Tunnel axis-oriented shield attitude precision control system and method

Abstract

The present invention discloses a shield attitude precision control system and method for tunnel axis target, the system includes: working condition data acquisition module, shield attitude optimization precision control module; the working condition data acquisition module is used to obtain the real-time construction parameters of the shield, read the attitude control target at the current position, extract the working condition characteristics of the current shield, and send the acquired information to the shield attitude optimization precision control module; the shield attitude optimization precision control module is used to optimize the jack zone oil pressure distribution of the propulsion system based on the ideal control target and the shield attitude auxiliary method based on machine learning technology, so as to realize the precise control of the shield attitude. The shield attitude precision control system and method for tunnel axis target disclosed by the present invention can realize the precise control of the shield attitude.

Description

Technical Field

[0001] The present invention belongs to the field of electronic information technology, and relates to a posture control system, and in particular to a shield posture precision control system and method facing a tunnel axis target. Background Art

[0002] The existing shield attitude control methods are divided into four categories:

[0003] (1) Qualitative empirical method: According to the relative position relationship between the shield machine's posture and the ideal control target, the shield machine's posture adjustment direction is determined, and based on this posture adjustment direction, the jack zone oil pressure on both sides is adjusted. This method completely relies on the shield machine driver's oil pressure adjustment experience, and adjusts the shield machine's posture through continuous trial and error. This method has low posture control accuracy and poor reliability of posture adjustment effect.

[0004] (2) Fuzzy control theory: Based on the principle of manual deviation correction, the input and output of the deviation correction controller are selected, and the fuzzy control rules for shield deviation correction are established according to the deviation correction experience of the shield driver. The change of the shield posture is not only affected by the current shield jack oil pressure distribution, but also by the external load on the shield. Comprehensively considering the influence of the jack oil pressure and the load on the shield will increase the design difficulty of the fuzzy controller and reduce the accuracy of the shield posture control. In addition, the design of the fuzzy control table is mostly based on the propulsion characteristics of a certain model of shield in a single tunnel. The generalization and portability of the model control table are poor. (Li Yueqiang, Chen Qingshan, Pan Zhikang. Design of automatic deviation correction control system for shield machine [J]. Journal of Beijing Information Science and Technology University: Natural Science Edition, 2014(29):10-15.)

[0005] (3) Classical control theory: Based on PID control theory, the oil pressure of each zone jack is adjusted according to the deviation between the shield machine posture and the ideal control target. Although the PID method has the advantages of fast response and stable control in control systems with analyzable input and output, it is very difficult to design and set parameters of PID for complex systems such as shield machine posture control, which are nonlinear, strongly coupled, and have large delays. (Liu Xiaonan, Ma Longhua. Research on deviation correction control of shield machine based on PSO-PID [J]. Computer Measurement and Control, 2020(9).)

[0006] (4) Machine learning method: In order to cope with the nonlinear, strong coupling and large delay characteristics of shield machine attitude control, some studies use machine learning methods to learn the kinematic laws of shield machines. However, such studies generally discuss the kinematic laws of shield machines under certain specific working conditions, and the models built have certain limitations. (Ding Haiying, Research on attitude and trajectory control of shield machines based on BP neural network [J]. Mechanical Design and Manufacturing Engineering, 2016(12):46-49.)

[0007] The above methods have gradually achieved certain theoretical research results on the problem of shield machine attitude control, but there are still certain defects, which can be summarized as follows: (1) The diversity of shield machine construction conditions during excavation is not considered; (2) The control ability of the shield machine propulsion system on the shield machine attitude is not considered; (3) The role of attitude control auxiliary methods in shield machine attitude control is not considered.

[0008] In view of this, there is an urgent need to design a new shield attitude control method to overcome at least some of the above-mentioned defects of the existing shield attitude control method. Summary of the invention

[0009] The present invention provides a tunnel axis target-oriented shield attitude precision control system and method, which can realize precise control of the shield attitude.

[0010] In order to solve the above technical problems, according to one aspect of the present invention, the following technical solution is adopted:

[0011] A shield attitude precision control system oriented to a tunnel axis target, the system comprising: a working condition data acquisition module, a working condition feature recognition module, and a shield attitude optimization precision control module;

[0012] The working condition data acquisition module is used to obtain the real-time construction parameters of the shield, read the attitude control target at the current position, extract the working condition characteristics of the current shield, and send the acquired information to the working condition characteristic recognition module and the shield attitude optimization precise control module;

[0013] The working condition feature recognition module is used to establish a working condition identifier based on massive historical shield tunnel construction data using a data-driven or rule-based method;

[0014] The shield attitude optimization and precise control module is used to extract shield construction data under the same working conditions from massive historical construction data, train to obtain a model of the relationship between the shield jack oil pressure distribution and attitude change, and optimize the jack zone oil pressure distribution of the propulsion system according to the ideal control target and the shield attitude auxiliary method to achieve precise control of the shield attitude.

[0015] As an implementation mode of the present invention, the working condition data acquisition module comprises: a tunnel construction parameter and shield tunneling parameter acquirer and a current working condition feature extractor;

[0016] The shield construction parameter acquirer is used to extract the current relevant construction parameters of the shield from the shield sensing, measurement, and posture planning system, and the extracted information includes at least the following dimensions: cutter head torque, propulsion speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, upper / right / lower / left partition jack oil pressure valve opening, articulation level, articulation height, shield posture control target, current shield posture, current jack grouping status, and current jack activation status; the shield construction parameter acquirer transmits the above parameters to the current working condition feature extractor and the shield posture precise optimization control module;

[0017] The current working condition feature extractor is used to extract the average value of the construction parameters of the shield during the latest construction ring excavation process from the shield construction parameter acquirer, which at least includes the following parameters: front soil pressure upper, front soil pressure right, front soil pressure lower, front soil pressure left, cutter head torque, cutter head speed, total thrust, oil pressure of the jacks in the upper / right / lower / left zones, as well as the degree of articulation level and the degree of articulation height.

[0018] As an implementation mode of the present invention, the working condition feature recognition module comprises a massive historical shield tunnel engineering construction data database and a working condition feature identifier;

[0019] The massive historical shield tunnel engineering construction data database is used to store massive historical shield tunnel engineering construction data, wherein the data dimensions at least include cutter head torque, propulsion speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, articulation horizontal degree, and articulation height degree;

[0020] The operating condition feature identifier is used to construct the operating condition feature identifier based on the operating condition feature data in massive historical projects. This process is established using a data-driven method or based on rules; the operating condition feature parameter type is consistent with the feature parameter type in the current operating condition feature extractor.

[0021] As an implementation mode of the present invention, the shield attitude optimization precise control module includes: a shield attitude control target construction module, a partitioned oil pressure attitude relationship model, a jack oil pressure valve opening setter, a shield attitude control capability evaluator, an attitude control auxiliary method library, and an attitude control parameter setting optimizer;

[0022] The shield attitude control target building module is used to obtain the control target of the current shield attitude from the shield attitude planning system, that is, the expected spatial change of the shield cutout and the shield tail in the horizontal elevation direction, and the change of the elevation inclination angle;

[0023] The partition oil pressure attitude relationship model is used to fit the relationship between the jack partition oil pressure and the shield attitude change under different working conditions according to historical data;

[0024] The jack oil pressure valve setter is used to adjust the corresponding oil pressure valve opening value according to the ideal oil pressure of each partition, and transmit the value to the oil pressure valve actuator;

[0025] The shield attitude control capability evaluator is used to evaluate whether the current jack zone oil pressure can control the shield to reach the ideal control target, and evaluate its control effect, and cooperate with the attitude control parameter setting optimizer to optimize the zone oil pressure distribution;

[0026] The attitude control parameter setting optimizer is used to adjust the oil pressure valve opening of the jack oil pressure valve opening setter in combination with the evaluation result of the shield attitude control effect evaluator, so as to adjust the oil pressure of each partition jack; when the attitude control parameter setting optimizer cannot find a feasible solution that meets the minimum index S of the shield attitude control effect evaluator by adjusting the oil pressure valve opening only, the relevant construction parameters of the current shield excavation are adjusted through the attitude control auxiliary method library; the attitude control parameter setting optimizer searches for the optimal jack partition oil pressure value distribution method according to the adjusted construction parameter configuration and the oil pressure attitude relationship model;

[0027] The posture control auxiliary method library is used to provide at least two posture control auxiliary methods, which adjust corresponding construction parameters when triggered and enabled, and assist the posture control parameter setting optimizer to find the optimal solution for achieving precise control of the shield posture.

[0028] As an implementation mode of the present invention, the posture control auxiliary method library includes a jack grouping adjustment method, a jack switch control method, and a shield articulation degree adjustment method;

[0029] When using the shield attitude control auxiliary method, the corresponding shield operation parameters are adjusted according to the method used, and the corresponding parameter adjustment status will be transmitted to the jack hydraulic valve opening setter.

[0030] The jack oil pressure valve opening setter is used to read the current upper / right / lower / left partition jack oil pressure and upper / right / lower / left partition jack oil pressure valve opening information from the shield construction parameter acquirer; read the relative position relationship between the current shield posture and the control target from the shield posture control target; read the current jack partition grouping method and the jack switch status from the posture control auxiliary method library; initialize the jack oil pressure valve opening value according to the above information. During the system operation, the jack oil pressure valve opening setter reads the jack oil pressure recommended value output by the PSO posture control parameter setting optimizer, updates the current shield each partition jack oil pressure valve opening setting value, and adjusts the current actual oil pressure value of each partition jack to the oil pressure recommended value;

[0031] The jack oil pressure valve opening setting device outputs the zone jack oil pressure adjustment result to a massive historical shield tunnel engineering construction data database.

[0032] As an embodiment of the present invention, the shield attitude optimization precise control module further includes a historical working condition feature identifier; the historical working condition feature extractor is used to cluster the historical working condition feature data based on massive historical engineering data, and its working condition feature parameter type is consistent with the feature parameter type in the current working condition feature extractor; the K-Means algorithm is used to cluster the historical working conditions, and the number of working condition categories is determined according to the following method:

[0033] Step 1: Assume that the initial number of working condition categories is k = 2, and k <n;

[0034] Step 2: Calculate the Euclidean distance D from each center point in the sample data when the number of categories is k using the K-Means method. k ;

[0035] Step 3: Add one to the number of current working conditions, i.e. k+1;

[0036] Step 4: Calculate the distance D between each sample and the center point after the number of working conditions increases by 1. k+1 , after the calculation category increases, the reduction in the Euclidean distance of the sample from the center point is G, that is, G = D k+1 -D k ;

[0037] Step 5: Repeat Step 2 to Step 4. When G decreases, the k value at this time is the number of categories that the historical working conditions should be divided into;

[0038] According to the above method, the number of categories determined by historical working conditions is determined, and their center points are numbered, which are used as working condition labels for historical construction data sections, and the working condition feature identifier is converted into a classifier function;

[0039] Read the current working condition feature parameters passed in by the current working condition feature extractor, judge the category to which the current working condition belongs based on the Euclidean distance, take the working condition type with the smallest Euclidean distance from each type as the category to which the current working condition belongs, and output the location stamps of all sections in this category.

[0040] The zoned oil pressure attitude relationship model is used to extract the zoned oil pressure of the jack and the corresponding shield attitude change of the corresponding section from the massive historical construction data based on the same working condition section position stamp output by the working condition feature identifier. Figure 2 The deep learning network shown establishes the relationship between the two; the input dimension of the network is the average oil pressure in the ε distance of the upper / right / lower / left partition jack, expressed as The output dimension is the spatial variation of the shield cut and shield tail in the horizontal elevation direction and the variation of the elevation inclination angle. The distance span of the variation is consistent with the distance span ε for calculating the average oil pressure of the jack, which is expressed as Among them, the calculation formula of hidden layer neuron data is:

[0041]

[0042] Among them, h j is the calculation result of hidden layer neurons; w ij is the weight value between the input layer neurons and the hidden layer neurons; θ j is the bias value between the input layer neurons and the hidden layer neurons; f is the activation function set for the hidden layer;

[0043] The zoned oil pressure attitude relationship model calculates the spatial position change of the shield attitude driven by the current jack oil pressure under the ε distance span according to the oil pressure value of each zone of the shield jack under the current valve opening input by the jack oil pressure valve opening setter.

[0044] The shield attitude control performance evaluator is used to evaluate the control effect based on the predicted value of the shield spatial position change at the ε distance and the ideal control target at the corresponding distance;

[0045] When the predicted attitude of the shield machine meets the requirements of the ideal control target, it means that the current jack zone oil pressure distribution meets the requirements of reaching the ideal control target; when the predicted result meets the ideal control target, it means that the current jack zone oil pressure distribution method cannot meet the requirements of reaching the ideal control target. The evaluation method of the current control capability of the shield machine should consider the following factors: (1) The control deviation D of each control indicator i; (2) Allowable error of each control index (3) The minimum tolerable index S of the overall fit.

[0046] The attitude control parameter setting optimizer is used to adjust the oil pressure valve opening of the jack oil pressure valve opening setter in combination with the evaluation result of the shield attitude control capability evaluator, so as to adjust the oil pressure of each partition jack; when the attitude control parameter setting optimizer cannot find a feasible solution that meets the minimum index S of the shield attitude control capability evaluator by adjusting the oil pressure valve opening only, the relevant construction parameters of the current shield excavation are adjusted through the attitude control auxiliary method library; the attitude control parameter setting optimizer searches for the optimal jack partition oil pressure value distribution method according to the adjusted construction parameter configuration and the oil pressure attitude relationship model;

[0047] The configuration of the attitude control parameter setting optimizer is as follows:

[0048] (1) The constraints of the attitude control parameter setting optimizer include: ① the minimum thrust for driving the shield tunneling; ② the maximum number of jacks allowed to be closed in each zone; ③ the minimum oil pressure required for a single group of jacks to extend; ④ the number of jack zones; ⑤ the number of jacks in each zone; ⑥ the maximum oil pressure of the propulsion system oil pump; ⑦ the oil pressure valve opening control range; ⑧ the oil pressure valve opening-oil pressure value characteristic curve relationship; ⑨ the jack zone grouping method;

[0049] (2) Evaluation function of the attitude control parameter setting optimizer: using the shield attitude control capability evaluator as the evaluation function in the optimization process;

[0050] (3) The thrust calculation formula of the partition jack of the attitude control parameter setting optimizer is: F i =Num i *f i ;

[0051] Among them, F i is the total pressure of the jack in zone i; Num i The number of jacks opened for partition i; f i The oil pressure of a single jack;

[0052] (4) Relevant configuration is performed on the optimization algorithm of the attitude control parameter setting optimizer.

[0053] According to another aspect of the present invention, the following technical solution is adopted: a shield attitude precision control method oriented to a tunnel axis target, the method comprising:

[0054] The working condition data acquisition step is to obtain the real-time construction parameters of the shield machine, read the attitude control target at the current position, and extract the working condition characteristics of the current shield machine;

[0055] The working condition feature recognition step is to establish a working condition identifier based on massive historical shield tunnel construction data using a data-driven or rule-based approach;

[0056] The precise control steps of shield attitude optimization are to extract shield construction data under the same working conditions from massive historical construction data, and train to obtain the relationship model between the shield jack oil pressure distribution and attitude change; according to the ideal control target and shield attitude auxiliary method, the jack zone oil pressure distribution of the propulsion system is optimized to achieve precise control of the shield attitude.

[0057] As an implementation mode of the present invention, the operating condition data acquisition step includes:

[0058] The shield construction parameter acquirer extracts the current relevant construction parameters of the shield from the shield sensor system, and the extracted information includes at least the following dimensions: cutter head torque, propulsion speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, upper / right / lower / left partition jack oil pressure valve opening, and articulation degree; the shield construction parameter acquirer transmits the above parameters to the current working condition feature extractor and the jack oil pressure valve opening setter;

[0059] The shield attitude control target building module obtains the control target of the current shield attitude from the shield trajectory planning system, that is, the expected spatial change of the shield cutout and the shield tail in the horizontal elevation direction, as well as the change of the elevation tilt angle; and transmits the control target to the jack hydraulic valve opening setter;

[0060] The current working condition feature extractor extracts the average value of the construction parameters of the shield during the latest construction ring excavation process from the shield construction parameter acquirer, which at least includes the following parameters: front soil pressure upper, front soil pressure right, front soil pressure lower, front soil pressure left, cutter head torque, cutter head speed, total thrust, oil pressure of the jacks in each upper / right / lower / left partition, and degree of articulation.

[0061] As an implementation mode of the present invention, the operating condition feature identification step includes:

[0062] Stores a large amount of historical shield tunnel construction data, where the data dimensions at least include cutter head torque, propulsion speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, articulation horizontal degree, and articulation height degree;

[0063] The operating condition feature identifier is constructed based on the operating condition feature data in massive historical projects. This process is established using a data-driven method or based on rules. The operating condition feature parameter type is consistent with the feature parameter type in the current operating condition feature extractor.

[0064] As an implementation mode of the present invention, the shield posture optimization and precise control step comprises:

[0065] A massive amount of historical shield tunnel construction data is stored in a massive historical shield tunnel construction data database, where the data dimensions of the tunnel construction data at least include cutter head torque, advancement speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, and articulation degree;

[0066] The historical working condition feature identifier clusters the excavation working conditions according to the historical construction parameters;

[0067] The zone oil pressure attitude relationship model fits the relationship between the zone oil pressure of the jack and the change of shield attitude under different working conditions according to historical data;

[0068] The jack hydraulic valve setter adjusts the corresponding hydraulic valve opening value according to the ideal hydraulic pressure in each zone, and transmits the value to the hydraulic valve actuator;

[0069] The shield attitude control capability evaluator evaluates whether the current jack zone oil pressure can control the shield to reach the ideal control target, and evaluates its control effect, and cooperates with the attitude control parameter setting optimizer to optimize the zone oil pressure distribution;

[0070] The attitude control parameter setting optimizer optimizes the oil pressure distribution of the jack zones in combination with the evaluation results of the shield attitude control capability evaluator, and hierarchically triggers the methods in the attitude control auxiliary method library according to the search results of the optimal solution;

[0071] The posture control auxiliary method library provides at least two posture control auxiliary methods. When triggered and enabled, it adjusts the corresponding construction parameters and assists the posture control parameter setting optimizer to find the optimal solution for achieving precise control of the shield posture.

[0072] As an implementation mode of the present invention, the posture control auxiliary method library includes a jack grouping adjustment method, a jack switch control method, and a shield articulation degree adjustment method; when the shield posture control auxiliary method is used, the corresponding shield operation parameters are adjusted accordingly according to the method used, and the corresponding parameter adjustment status will be transmitted to the jack oil pressure valve opening setter;

[0073] The jack oil pressure valve opening setter reads the current upper / right / lower / left partition jack oil pressure and upper / right / lower / left partition jack oil pressure valve opening information from the shield construction parameter acquirer; reads the relative position relationship between the current shield posture and the control target from the shield posture control target; reads the current jack partition grouping method and the jack switch status from the posture control auxiliary method library; initializes the jack oil pressure valve opening value according to the above information. During the system operation, the jack oil pressure valve opening setter reads the jack oil pressure recommended value output by the PSO posture control parameter setting optimizer, updates the current shield partition jack oil pressure valve opening setting value, and adjusts the current actual oil pressure value of the jack in each partition to the oil pressure recommended value;

[0074] The jack oil pressure valve opening setting device outputs the zone jack oil pressure adjustment result to a massive historical shield tunnel engineering construction data database.

[0075] As an implementation mode of the present invention, the shield posture optimization and precise control step includes a historical working condition feature extraction step, including: clustering the historical working condition feature data based on massive historical engineering data by a historical working condition feature extractor, and the working condition feature parameter type is consistent with the feature parameter type in the current working condition feature extractor; clustering the historical working conditions using a K-Means algorithm, and the number of working condition categories is determined according to the following method:

[0076] Step 1: Assume that the initial number of working condition categories is k = 2, and k <n;

[0077] Step 2: Calculate the Euclidean distance D from each center point in the sample data when the number of categories is k using the K-Means method. k ;

[0078] Step 3: Add one to the number of current working conditions, i.e. k+1;

[0079] Step 4: Calculate the distance D between each sample and the center point after the number of working conditions increases by 1. k+1 , after the calculation category increases, the reduction in the Euclidean distance of the sample from the center point is G, that is, G = D k+1 -D k ;

[0080] Step 5: Repeat Step 2 to Step 4. When G decreases, the k value at this time is the number of categories that the historical working conditions should be divided into;

[0081] According to the above method, the number of categories determined by historical working conditions is determined, and their center points are numbered, which are used as working condition labels for historical construction data sections, and the working condition feature identifier is converted into a classifier function;

[0082] Read the current working condition feature parameters passed in by the current working condition feature extractor, judge the category to which the current working condition belongs based on the Euclidean distance, take the working condition type with the smallest Euclidean distance from each type as the category to which the current working condition belongs, and output the location stamps of all sections in this category.

[0083] The zoned oil pressure attitude relationship model is based on the same working condition section position stamp output by the working condition feature identifier, and extracts the corresponding section jack zone oil pressure and the corresponding shield attitude change from the massive historical construction data. Figure 2 The deep learning network shown establishes the relationship between the two; the input dimension of the network is the average oil pressure in the ε distance of the upper / right / lower / left partition jack, expressed as The output dimension is the spatial variation of the shield cut and shield tail in the horizontal elevation direction and the variation of the elevation inclination angle. The distance span of the variation is consistent with the distance span ε for calculating the average oil pressure of the jack, which is expressed as Among them, the calculation formula of hidden layer neuron data is:

[0084]

[0085] Among them, h j is the calculation result of hidden layer neurons; w ij is the weight value between the input layer neurons and the hidden layer neurons; θ j is the bias value between the input layer neurons and the hidden layer neurons; f is the activation function set for the hidden layer;

[0086] The zoned oil pressure attitude relationship model calculates the spatial position change of the shield attitude driven by the current jack oil pressure under the ε distance span according to the oil pressure value of each zone of the shield jack under the current valve opening input by the jack oil pressure valve opening setter.

[0087] The shield attitude control performance evaluator evaluates the control effect based on the predicted value of the shield spatial position change at the ε distance and the ideal control target at the corresponding distance;

[0088] When the predicted attitude of the shield machine meets the requirements of the ideal control target, it means that the current jack zone oil pressure distribution meets the requirements of reaching the ideal control target; when the predicted result meets the ideal control target, it means that the current jack zone oil pressure distribution method cannot meet the requirements of reaching the ideal control target. The evaluation method of the current control capability of the shield machine should consider the following factors: (1) The control deviation D of each control indicator i ; (2) Allowable error of each control index (3) The minimum tolerable index S of the overall fit.

[0089] The attitude control parameter setting optimizer adjusts the oil pressure valve opening of the jack oil pressure valve opening setter in combination with the evaluation result of the shield attitude control capability evaluator, thereby adjusting the oil pressure of each partition jack; when the attitude control parameter setting optimizer cannot find a feasible solution that meets the minimum index S of the shield attitude control capability evaluator by adjusting the oil pressure valve opening only, the relevant construction parameters of the current shield excavation are adjusted through the attitude control auxiliary method library; the attitude control parameter setting optimizer searches for the optimal jack partition oil pressure value distribution method according to the adjusted construction parameter configuration and the oil pressure attitude relationship model;

[0090] The configuration of the attitude control parameter setting optimizer is as follows:

[0091] 1) The constraints of the attitude control parameter setting optimizer include: ① the minimum thrust for driving the shield tunneling; ② the maximum number of jacks allowed to be closed in each partition; ③ the minimum oil pressure value required for a single group of jacks to extend; ④ the number of jack partitions; ⑤ the number of jacks in each partition; ⑥ the maximum oil pressure value of the propulsion system oil pump; ⑦ the oil pressure valve opening control range; ⑧ the oil pressure valve opening-oil pressure value characteristic curve relationship; ⑨ the jack partition grouping method;

[0092] 2) Evaluation function of the attitude control parameter setting optimizer: using the shield attitude control capability evaluator as the evaluation function in the optimization process;

[0093] 3) The partition jack thrust calculation formula of the attitude control parameter setting optimizer is: F i =Num i *f i ;

[0094] Among them, F i is the total pressure of the jack in zone i; Num i The number of jacks opened for partition i; f i The oil pressure of a single jack;

[0095] 4) Relevant configuration is performed on the optimization algorithm of the attitude control parameter setting optimizer.

[0096] The beneficial effect of the present invention is that the shield attitude precision control system and method oriented to the tunnel axis target proposed in the present invention can realize precise control of the shield attitude. BRIEF DESCRIPTION OF THE DRAWINGS

[0097] Figure 1 It is a schematic diagram of the composition of a shield machine posture precision control system in one embodiment of the present invention.

[0098] Figure 2 It is a network structure diagram of the zoned oil pressure posture relationship model in one embodiment of the present invention.

[0099] Figure 3 This is a flow chart of optimizing the oil pressure distribution of jack zones based on PSO in one embodiment of the present invention. DETAILED DESCRIPTION

[0100] The preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0101] In order to further understand the present invention, preferred embodiments of the present invention are described below in conjunction with examples. However, it should be understood that these descriptions are only for further illustrating the features and advantages of the present invention, rather than limiting the claims of the present invention.

[0102] The description in this section is only for several typical embodiments, and the present invention is not limited to the scope of the embodiments. The same or similar prior art means and some technical features in the embodiments are mutually replaced within the scope of the present invention.

[0103] The description of the steps in each embodiment in the specification is only for the convenience of explanation, and the implementation method of this application is not limited by the order of step implementation. The "connection" in the specification includes both direct connection and indirect connection. The concepts of up / right / down / left mentioned in the specification are described for the convenience of understanding a certain partitioning method selected, which also includes other forms of jack partitioning; the shield posture mentioned in the patent of this invention is any expression method to describe the spatial position of the shield. In order to better introduce how to use this patent, use To explain, but This is a description method for describing the shield posture, and the scope of this patent includes similar description methods.

[0104] The present invention discloses a shield attitude precision control system oriented to the tunnel axis target. Figure 1 This is a schematic diagram of the components of a shield attitude precision control system in one embodiment of the present invention; please refer to Figure 1 The shield attitude precise control system includes: a working condition data acquisition module 1, a working condition feature recognition module 2 and a shield attitude optimization precise control module 3.

[0105] The working condition data acquisition module 1 is used to obtain the real-time construction parameters of the shield, read the attitude control target at the current position, extract the working condition characteristics of the current shield, and send the acquired information to the working condition characteristic recognition module 2 and the shield attitude optimization precise control module 3.

[0106] The working condition feature identification module 2 is used to establish a working condition identifier based on massive historical shield tunnel construction data using a data-driven or rule-based method.

[0107] The shield attitude optimization precise control module 3 is used to optimize the jack zone oil pressure distribution of the propulsion system based on the machine learning technology according to the ideal control target and the shield attitude auxiliary method to achieve precise control of the shield attitude.

[0108] In one embodiment of the present invention, the working condition data acquisition module 1 includes: a tunnel construction parameter and shield tunneling parameter acquirer 11 and a current working condition feature extractor 13 .

[0109] The shield construction parameter acquirer 11 is used to extract the current relevant construction parameters of the shield from the shield sensing, measurement, and posture planning systems. The extracted information includes at least the following dimensions: cutter head torque, propulsion speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, upper / right / lower / left partition jack oil pressure valve opening, articulation level, articulation height, shield posture control target, current shield posture, current jack grouping status, and current jack activation status; the shield construction parameter acquirer transmits the above parameters to the current working condition feature extractor 13 and the shield posture precise optimization control module 3.

[0110] The current working condition feature extractor 13 is used to extract the average value of the construction parameters of the shield during the latest construction ring excavation process from the shield construction parameter acquirer, which at least includes the following parameters: front soil pressure upper, front soil pressure right, front soil pressure lower, front soil pressure left, cutter head torque, cutter head speed, total thrust, oil pressure of the jacks in the upper / right / lower / left zones, and articulation horizontal degree and articulation height degree.

[0111] The working condition feature recognition module 2 includes a massive historical shield tunnel engineering construction data database 21 and a working condition feature identifier 22 .

[0112] The massive historical shield tunnel engineering construction data database 21 is used to store massive historical shield tunnel engineering construction data, wherein the data dimensions at least include cutter head torque, propulsion speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left zone jack oil pressure, articulation horizontal degree, and articulation height degree.

[0113] The operating condition feature identifier 22 is used to construct an operating condition feature identifier based on operating condition feature data in massive historical projects. This process is established using a data-driven method or based on rules; the operating condition feature parameter type is consistent with the feature parameter type in the current operating condition feature extractor.

[0114] In one embodiment of the present invention, the shield attitude optimization precise control module 3 includes: a shield attitude control target construction module 31, a partitioned oil pressure attitude relationship model 32, a jack oil pressure valve opening setter 35, a shield attitude control capability evaluator 33, an attitude control auxiliary method library 36, and an attitude control parameter setting optimizer 34. In addition, the jack oil pressure valve opening setter 35 and the attitude control auxiliary method library 36 are respectively connected to the shield machine 37.

[0115] The shield attitude control target construction module 31 is used to obtain the control target of the current shield attitude from the shield attitude planning system, that is, the expected spatial variation of the shield cutout and the shield tail in the horizontal elevation direction, and the variation of the elevation tilt angle.

[0116] The partitioned oil pressure attitude relationship model 32 is used to fit the relationship between the partitioned oil pressure of the jack and the change in shield attitude under different working conditions according to historical data.

[0117] The jack oil pressure valve setter 25 is used to adjust the corresponding oil pressure valve opening value according to the ideal oil pressure of each partition, and transmit the value to the oil pressure valve actuator.

[0118] The shield attitude control capability evaluator 33 is used to evaluate whether the current jack zone oil pressure can control the shield to reach the ideal control target, and evaluate its control effect, and cooperate with the attitude control parameter setting optimizer to optimize the zone oil pressure distribution.

[0119] The attitude control parameter setting optimizer 34 is used to optimize the jack zone oil pressure distribution in combination with the evaluation result of the shield attitude control capability evaluator, and hierarchically trigger the methods in the attitude control auxiliary method library 26 according to the search result of the optimal solution.

[0120] In one embodiment, the attitude control parameter setting optimizer 34 is used to adjust the oil pressure valve opening of the jack oil pressure valve opening setter in combination with the evaluation result of the shield attitude control effect evaluator, thereby adjusting the oil pressure of each partition jack; when the attitude control parameter setting optimizer cannot find a feasible solution that meets the minimum index S of the shield attitude control effect evaluator by simply adjusting the oil pressure valve opening, the relevant construction parameters of the current shield excavation are adjusted through the attitude control auxiliary method library; the attitude control parameter setting optimizer searches for the optimal jack partition oil pressure value distribution method based on the adjusted construction parameter configuration and the oil pressure attitude relationship model.

[0121] The posture control auxiliary method library 36 is used to provide at least two posture control auxiliary methods, which adjust corresponding construction parameters when triggered and activated, and assist the posture control parameter setting optimizer to find the optimal solution for achieving precise control of the shield posture.

[0122] In one embodiment of the present invention, the posture control auxiliary method library 36 includes a jack group adjustment method, a jack switch control method, and a shield articulation degree adjustment method; when the shield posture control auxiliary method is used, the corresponding shield operation parameters are adjusted accordingly according to the method used, and the corresponding parameter adjustment situation will be transmitted to the jack oil pressure valve opening setter 24.

[0123] The jack oil pressure valve opening setter 35 is used to read the current upper / right / lower / left partition jack oil pressure and upper / right / lower / left partition jack oil pressure valve opening information from the shield construction parameter acquirer; read the relative position relationship between the current shield posture and the control target from the shield posture control target; read the current jack partition grouping method and the jack switch status from the posture control auxiliary method library; initialize the jack oil pressure valve opening value according to the above information. During the system operation, the jack oil pressure valve opening setter reads the jack oil pressure recommended value output by the PSO posture control parameter setting optimizer, updates the current shield each partition jack oil pressure valve opening setting value, and adjusts the current actual oil pressure value of each partition jack to the oil pressure recommended value;

[0124] The jack oil pressure valve opening setter 35 outputs the zone jack oil pressure adjustment result to a massive historical shield tunnel engineering construction data database.

[0125] In one embodiment of the present invention, the shield attitude optimization precise control module further includes a historical working condition feature identifier. The historical working condition feature extractor is used to cluster the historical working condition feature data based on massive historical engineering data, and its working condition feature parameter type is consistent with the feature parameter type in the current working condition feature extractor; the K-Means algorithm is used to cluster the historical working conditions, and the number of working condition categories is determined according to the following method:

[0126] Step 1: Assume that the initial number of working condition categories is k = 2, and k <n;

[0127] Step 2: Calculate the Euclidean distance D from each center point in the sample data when the number of categories is k using the K-Means method. k ;

[0128] Step 3: Add one to the number of current working conditions, i.e. k+1;

[0129] Step 4: Calculate the distance D between each sample and the center point after the number of working conditions increases by 1. k+1 , after the calculation category increases, the reduction in the Euclidean distance of the sample from the center point is G, that is, G = D k+1 -D k ;

[0130] Step 5: Repeat Step 2 to Step 4. When G decreases, the k value at this time is the number of categories that the historical working conditions should be divided into;

[0131] According to the above method, the number of categories determined by historical working conditions is determined, and their center points are numbered, which are used as working condition labels for historical construction data sections, and the working condition feature identifier is converted into a classifier function;

[0132] Read the current operating condition feature parameters passed in by the current operating condition feature extractor 13, judge the category to which the current operating condition belongs based on the Euclidean distance, take the operating condition type with the smallest Euclidean distance from each type as the category to which the current operating condition belongs, and output the location stamps of all sections in this category.

[0133] Figure 2 This is a network structure diagram of the zoned oil pressure posture relationship model in one embodiment of the present invention; please refer to Figure 2 In one embodiment of the present invention, the zoned oil pressure attitude relationship model 32 is used to extract the zoned oil pressure of the jack and the corresponding shield attitude change of the corresponding section from the massive historical construction data based on the same working condition section position stamp output by the working condition feature identifier, Figure 2 The deep learning network shown establishes the relationship between the two; the input dimension of the network is the average oil pressure in the ε distance of the upper / right / lower / left partition jack, expressed as The output dimension is the spatial variation of the shield cut and shield tail in the horizontal elevation direction and the variation of the elevation inclination angle. The distance span of the variation is consistent with the distance span ε for calculating the average oil pressure of the jack, which is expressed as Among them, the calculation formula of hidden layer neuron data is:

[0134]

[0135] Among them, h j is the calculation result of hidden layer neurons; w ij is the weight value between the input layer neurons and the hidden layer neurons; θ j is the bias value between the input layer neurons and the hidden layer neurons; f is the activation function set for the hidden layer;

[0136] The zone oil pressure attitude relationship model 32 calculates the spatial position change of the shield attitude driven by the current jack oil pressure at the ε distance span according to the oil pressure value of each zone of the shield jack under the current valve opening input by the jack oil pressure valve opening setter.

[0137] In one embodiment of the present invention, the shield attitude control performance evaluator 33 is used to evaluate the control effect based on the predicted value of the shield spatial position change at the ε distance and the ideal control target at the corresponding distance;

[0138] When the predicted attitude of the shield machine meets the requirements of the ideal control target, it means that the current jack zone oil pressure distribution meets the requirements of reaching the ideal control target; when the predicted result meets the ideal control target, it means that the current jack zone oil pressure distribution method cannot meet the requirements of reaching the ideal control target. The evaluation method of the current control capability of the shield machine should consider the following factors: (1) The control deviation D of each control indicator i ; (2) Allowable error of each control index (3) The minimum tolerable index S of the overall fit.

[0139] The attitude control parameter setting optimizer 34 is used to adjust the oil pressure valve opening of the jack oil pressure valve opening setter in combination with the evaluation result of the shield attitude control capability evaluator, so as to adjust the oil pressure of each partition jack; when the attitude control parameter setting optimizer cannot find a feasible solution that meets the minimum index S of the shield attitude control capability evaluator by adjusting the oil pressure valve opening only, the relevant construction parameters of the current shield excavation are adjusted through the attitude control auxiliary method library; the attitude control parameter setting optimizer searches for the optimal jack partition oil pressure value distribution method according to the adjusted construction parameter configuration and the oil pressure attitude relationship model;

[0140] The configuration of the attitude control parameter setting optimizer 34 is as follows:

[0141] (1) The constraints of the attitude control parameter setting optimizer include: ① the minimum thrust for driving the shield tunneling; ② the maximum number of jacks allowed to be closed in each zone; ③ the minimum oil pressure required for a single group of jacks to extend; ④ the number of jack zones; ⑤ the number of jacks in each zone; ⑥ the maximum oil pressure of the propulsion system oil pump; ⑦ the oil pressure valve opening control range; ⑧ the oil pressure valve opening-oil pressure value characteristic curve relationship; ⑨ the jack zone grouping method;

[0142] (2) Evaluation function of the attitude control parameter setting optimizer: using the shield attitude control capability evaluator as the evaluation function in the optimization process;

[0143] (3) The thrust calculation formula of the partition jack of the attitude control parameter setting optimizer is: F i =Num i *f i ;

[0144] Among them, F i is the total pressure of the jack in zone i; Num i The number of jacks opened for partition i; f i The oil pressure of a single jack;

[0145] (4) The posture control parameter setting optimizer 27 has an optimization algorithm configuration.

[0146] Figure 3 This is a flowchart of optimizing the oil pressure distribution of jack zones based on PSO in one embodiment of the present invention; please refer to Figure 3 In one embodiment, the working process of the attitude control parameter setting optimizer is understood, and the PSO algorithm is selected here to further illustrate it.

[0147] The number of particles in the particle swarm;

[0148] The maximum number of iterations under the current construction parameter setting conditions;

[0149] The following formula is the iterative calculation process of a single particle in a particle swarm:

[0150] v i+1 =ω×v i +c 1 ×rand()×(pbest i -x i )+c 2 ×rand()×(gbest i -x i ) (1)

[0151] x i+1 =x i +v i+1 (2)

[0152] o i =f(x i ) (3)

[0153] In formula (1), (2), and (3), i represents the i-th attempt to set the oil pressure distribution of the jack partition; rand() represents a random number between (0, 1); c 1 and c 2 is the learning factor; v i Represents the adjustment direction of the current jack zone oil pressure distribution; x i Represents the current setting of the oil pressure of each zone of the jack; o i The jack zone valve opening is obtained based on the oil pressure valve opening-oil pressure value characteristic curve relationship f() of the current zone oil pressure.

[0154] The present invention also discloses a shield attitude precision control method facing a tunnel axis target, the shield attitude precision control method comprising:

[0155] The working condition data acquisition step is to acquire the real-time construction parameters of the shield, read the attitude control target at the current position, extract the working condition characteristics of the current shield, and send the acquired information to the shield attitude optimization precise control module;

[0156] The precise control steps of shield attitude optimization are based on machine learning technology and optimize the jack zone oil pressure distribution of the propulsion system according to the ideal control target and shield attitude auxiliary method to achieve precise control of the shield attitude.

[0157] In one embodiment of the present invention, the operating condition data acquisition step includes:

[0158] The shield construction parameter acquirer extracts the current relevant construction parameters of the shield from the shield sensor system, and the extracted information includes at least the following dimensions: cutter head torque, propulsion speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, upper / right / lower / left partition jack oil pressure valve opening, and articulation degree; the shield construction parameter acquirer transmits the above parameters to the current working condition feature extractor and the jack oil pressure valve opening setter;

[0159] The shield attitude control target building module obtains the control target of the current shield attitude from the shield trajectory planning system, that is, the expected spatial change of the shield cutout and the shield tail in the horizontal elevation direction, as well as the change of the elevation tilt angle; and transmits the control target to the jack hydraulic valve opening setter;

[0160] The current working condition feature extractor extracts the average value of the construction parameters of the shield during the latest construction ring excavation process from the shield construction parameter acquirer, which at least includes the following parameters: front soil pressure upper, front soil pressure right, front soil pressure lower, front soil pressure left, cutter head torque, cutter head speed, total thrust, oil pressure of the jacks in each upper / right / lower / left partition, and degree of articulation.

[0161] In one embodiment of the present invention, the shield posture optimization and precise control step comprises:

[0162] A massive amount of historical shield tunnel construction data is stored in a massive historical shield tunnel construction data database, where the data dimensions of the tunnel construction data at least include cutter head torque, advancement speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, and articulation degree;

[0163] The historical working condition feature identifier clusters the excavation working conditions according to the historical construction parameters;

[0164] The zone oil pressure attitude relationship model fits the relationship between the zone oil pressure of the jack and the change of shield attitude under different working conditions according to historical data;

[0165] The jack hydraulic valve setter adjusts the corresponding hydraulic valve opening value according to the ideal hydraulic pressure in each zone, and transmits the value to the hydraulic valve actuator;

[0166] The shield attitude control capability evaluator evaluates whether the current jack zone oil pressure can control the shield to reach the ideal control target, and evaluates its control effect, and cooperates with the attitude control parameter setting optimizer to optimize the zone oil pressure distribution;

[0167] The attitude control parameter setting optimizer optimizes the oil pressure distribution of the jack zones in combination with the evaluation results of the shield attitude control capability evaluator, and hierarchically triggers the methods in the attitude control auxiliary method library according to the search results of the optimal solution;

[0168] The posture control auxiliary method library provides at least two posture control auxiliary methods. When triggered and enabled, it adjusts the corresponding construction parameters and assists the posture control parameter setting optimizer to find the optimal solution for achieving precise control of the shield posture.

[0169] In one embodiment of the present invention, the posture control auxiliary method library includes the adjustment method of the jack grouping, the jack switch control method, and the adjustment method of the shield articulation degree; when the shield posture control auxiliary method is used, the corresponding shield operation parameters are adjusted accordingly according to the method used, and the corresponding parameter adjustment situation will be transmitted to the jack oil pressure valve opening setter;

[0170] The jack oil pressure valve opening setter reads the current upper / right / lower / left partition jack oil pressure and upper / right / lower / left partition jack oil pressure valve opening information from the shield construction parameter acquirer; reads the relative position relationship between the current shield posture and the control target from the shield posture control target; reads the current jack partition grouping method and jack switch status from the posture control auxiliary method library; initializes the jack oil pressure valve opening value according to the above information. During the system operation, the jack oil pressure valve opening setter reads the jack oil pressure recommended value output by the PSO posture control parameter setting optimizer, updates the current shield each partition jack oil pressure valve opening setting value, and adjusts the current actual oil pressure value of each partition jack to the oil pressure recommended value. The jack oil pressure valve opening setter outputs the partition jack oil pressure adjustment result to the massive historical shield tunnel engineering construction data database.

[0171] In one embodiment of the present invention, the historical operating condition feature extractor clusters the historical operating condition feature data based on the massive historical engineering data, and the operating condition feature parameter type is consistent with the feature parameter type in the current operating condition feature extractor; the K-Means algorithm is used to cluster the historical operating conditions, and the number of operating condition categories is determined according to the following method:

[0172] Step 1: Assume that the initial number of working condition categories is k = 2, and k <n;

[0173] Step 2: Calculate the Euclidean distance D from each center point in the sample data when the number of categories is k using the K-Means method. k ;

[0174] Step 3: Add one to the number of current working conditions, i.e. k+1;

[0175] Step 4: Calculate the distance D between each sample and the center point after the number of working conditions increases by 1. k+1 , after the calculation category increases, the reduction in the Euclidean distance of the sample from the center point is G, that is, G = D k+1 -D k ;

[0176] Step 5: Repeat Step 2 to Step 4. When G decreases, the k value at this time is the number of categories that the historical working conditions should be divided into;

[0177] According to the above method, the number of categories determined by historical working conditions is determined, and their center points are numbered, which are used as working condition labels for historical construction data sections, and the working condition feature identifier is converted into a classifier function;

[0178] Read the current working condition feature parameters passed in by the current working condition feature extractor, judge the category to which the current working condition belongs based on the Euclidean distance, take the working condition type with the smallest Euclidean distance from each type as the category to which the current working condition belongs, and output the location stamps of all sections in this category.

[0179] In one embodiment of the present invention, the zoned oil pressure attitude relationship model is based on the same working condition section position stamp output by the working condition feature identifier, and extracts the jack zoned oil pressure and the corresponding shield attitude change of the corresponding section from the massive historical construction data. Figure 2 The deep learning network shown establishes the relationship between the two; the input dimension of the network is the average oil pressure in the ε distance of the upper / right / lower / left partition jack, expressed as The output dimension is the spatial variation of the shield cut and shield tail in the horizontal elevation direction and the variation of the elevation inclination angle. The distance span of the variation is consistent with the distance span ε for calculating the average oil pressure of the jack, which is expressed as Among them, the calculation formula of hidden layer neuron data is:

[0180]

[0181] Among them, h j is the calculation result of hidden layer neurons; w ijis the weight value between the input layer neurons and the hidden layer neurons; θ j is the bias value between the input layer neurons and the hidden layer neurons; f is the activation function set for the hidden layer;

[0182] The zone oil pressure attitude relationship model calculates the spatial position change of the shield attitude driven by the current jack oil pressure at the ε distance span according to the oil pressure value of each zone of the shield jack under the current valve opening value input by the jack oil pressure valve opening setter. The ε value can be determined according to the engineering characteristics, and the recommended value range is 5mm to 10mm.

[0183] In one embodiment of the present invention, the shield attitude control performance evaluator evaluates the control effect based on the predicted value of the shield spatial position change at the ε distance and the ideal control target at the corresponding distance.

[0184] When the predicted attitude of the shield machine meets the requirements of the ideal control target, it means that the current jack zone oil pressure distribution meets the requirements of reaching the ideal control target; when the predicted result meets the ideal control target, it means that the current jack zone oil pressure distribution method cannot meet the requirements of reaching the ideal control target. The evaluation method of the current control capability of the shield machine should consider the following factors: (1) The control deviation D of each control indicator i ; (2) Allowable error of each control index (3) The minimum tolerable index S of the overall fit.

[0185] In one embodiment of the present invention, the attitude control parameter setting optimizer adjusts the oil pressure valve opening of the jack oil pressure valve opening setter in combination with the evaluation result of the shield attitude control capability evaluator, thereby adjusting the oil pressure of each partition jack; when the attitude control parameter setting optimizer cannot find a feasible solution that meets the minimum index S of the shield attitude control capability evaluator by adjusting the oil pressure valve opening only, the attitude control auxiliary method library is used to adjust the relevant construction parameters of the current shield excavation; the attitude control parameter setting optimizer searches for the optimal jack partition oil pressure value distribution method according to the adjusted construction parameter configuration and the oil pressure attitude relationship model;

[0186] The configuration of the attitude control parameter setting optimizer is as follows:

[0187] (1) The constraints of the attitude control parameter setting optimizer include: ① the minimum thrust for driving the shield tunneling; ② the maximum number of jacks allowed to be closed in each zone; ③ the minimum oil pressure required for a single group of jacks to extend; ④ the number of jack zones; ⑤ the number of jacks in each zone; ⑥ the maximum oil pressure of the propulsion system oil pump; ⑦ the oil pressure valve opening control range; ⑧ the oil pressure valve opening-oil pressure value characteristic curve relationship; ⑨ the jack zone grouping method;

[0188] (2) Evaluation function of the attitude control parameter setting optimizer: using the shield attitude control capability evaluator as the evaluation function in the optimization process;

[0189] In one embodiment, the evaluation function may be calculated as follows:

[0190]

[0191] Among them, Devi j is the deviation value of the shield type j posture description index, It is the expected maximum deviation value of the corresponding j-type attitude index, which can be set according to the expected control accuracy and engineering control requirements. When score ≥ η, it means that the current shield control performance has changed due to the change of working conditions, and the current working condition characteristics are extracted; when score < η, the current working condition is relatively stable, and the shield control performance is basically stable without major changes; the value range of η is recommended to be set to 10% to 20%.

[0192] (3) The thrust calculation formula of the partition jack of the attitude control parameter setting optimizer is: F i =Num i *f i ;

[0193] Among them, F i is the total pressure of the jack in zone i; Num i The number of jacks opened for partition i; f i The oil pressure of a single jack;

[0194] 4) Relevant configuration is performed on the optimization algorithm of the attitude control parameter setting optimizer:

[0195] Figure 3 This is a flowchart of optimizing the oil pressure distribution of jack zones based on PSO in one embodiment of the present invention; please refer to Figure 3 In one embodiment, the working process of the attitude control parameter setting optimizer uses the PSO algorithm;

[0196] Get the number of particles in the particle swarm;

[0197] Calculate the maximum number of iterations under the current construction parameter setting conditions; the following formula is the iterative calculation process of a single particle in the particle swarm:

[0198] v i+1 =ω×v i +c 1 ×rand()×(pbest i -x i )+c 2 ×rand()×(gbesti -x i ) (1)

[0199] x i+1 =x i +v i+1 (2)

[0200] o i =f(x i ) (3)

[0201] In formula (1), (2), and (3), i represents the i-th attempt to set the oil pressure distribution of the jack partition; rand() represents a random number between (0, 1); c 1 and c 2 is the learning factor; v i Represents the adjustment direction of the current jack zone oil pressure distribution; x i Represents the current setting of the oil pressure of each zone of the jack; o i The jack zone valve opening is obtained based on the oil pressure valve opening-oil pressure value characteristic curve relationship f() of the current zone oil pressure.

[0202] In summary, the precise control system and method of shield attitude oriented to the tunnel axis target can realize precise control of the shield attitude.

[0203] It should be noted that the present application can be implemented in software and / or a combination of software and hardware; for example, it can be implemented using an application-specific integrated circuit (ASIC), a general-purpose computer, or any other similar hardware device. In some embodiments, the software program of the present application can be executed by a processor to implement the above steps or functions. Similarly, the software program of the present application (including related data structures) can be stored in a computer-readable recording medium; for example, a RAM memory, a magnetic or optical drive or a floppy disk and the like. In addition, some steps or functions of the present application can be implemented in hardware; for example, as a circuit that cooperates with a processor to perform various steps or functions.

[0204] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0205] The description and application of the present invention here are illustrative, and it is not intended to limit the scope of the present invention to the above-mentioned embodiments. The effects or advantages involved in the embodiments may not be embodied in the embodiments due to interference from various factors, and the description of the effects or advantages is not used to limit the embodiments. The deformation and change of the embodiments disclosed here are possible, and the replacement of the embodiments and the various equivalent parts are well known to those of ordinary skill in the art. It should be clear to those skilled in the art that the present invention can be implemented in other forms, structures, arrangements, proportions, and with other components, materials and parts without departing from the spirit or essential features of the present invention. Other deformations and changes can be made to the embodiments disclosed here without departing from the scope and spirit of the present invention.

Claims

1. A shield attitude precision control system oriented to the tunnel axis target, characterized in that: The system includes: a working condition data acquisition module, a working condition feature recognition module, and a shield attitude optimization and precise control module; The working condition data acquisition module is used to obtain the real-time construction parameters and excavation parameters of the shield, extract the working condition characteristics of the current shield, and send the acquired information to the working condition characteristic recognition module and the shield posture optimization precise control module; The working condition feature recognition module is used to establish a working condition identifier based on massive historical shield tunnel construction data using a data-driven or rule-based method; The shield attitude optimization and precise control module is used to extract shield construction data under the same working conditions from massive historical construction data, train and obtain the relationship model between the shield jack oil pressure distribution and attitude change, and optimize the jack zone oil pressure distribution of the propulsion system according to the ideal control target and the shield attitude auxiliary method to achieve precise control of the shield attitude; The shield attitude optimization and precise control module includes: a shield attitude control target construction module, a partitioned oil pressure attitude relationship model, a jack oil pressure valve opening setter, a shield attitude control capability evaluator, an attitude control auxiliary method library, and an attitude control parameter setting optimizer; The shield attitude control target construction module is used to obtain the control target of the current shield attitude from the shield attitude planning system, that is, the expected spatial change of the shield cutout and the shield tail in the horizontal elevation direction, and the change of the elevation inclination angle; the partition oil pressure attitude relationship model is used to fit the relationship between the jack partition oil pressure and the shield attitude change under different working conditions according to historical data; The jack oil pressure valve opening setting device is used to adjust the corresponding oil pressure valve opening value according to the ideal oil pressure of each partition, and transmit the value to the oil pressure valve actuator; The shield attitude control capability evaluator is used to evaluate whether the current jack zone oil pressure can control the shield to reach the ideal control target, and evaluate its control effect, and cooperate with the attitude control parameter setting optimizer to optimize the zone oil pressure distribution; The attitude control parameter setting optimizer is used to adjust the oil pressure valve opening of the jack oil pressure valve opening setter in combination with the evaluation result of the shield attitude control capability evaluator, so as to adjust the oil pressure of each partition jack; when the attitude control parameter setting optimizer cannot find a feasible solution that meets the minimum index S of the shield attitude control capability evaluator by adjusting the oil pressure valve opening only, the relevant construction parameters of the current shield excavation are adjusted through the attitude control auxiliary method library; the attitude control parameter setting optimizer searches for the optimal jack partition oil pressure value distribution method according to the adjusted construction parameter configuration and the oil pressure attitude relationship model; The posture control auxiliary method library is used to provide at least two posture control auxiliary methods, which adjust corresponding construction parameters when triggered and enabled, and assist the posture control parameter setting optimizer to find the optimal solution for achieving precise control of shield posture; The posture control auxiliary method library includes the adjustment method of jack grouping, the jack switch control method, and the adjustment method of shield articulation degree; when the shield posture control auxiliary method is used, the corresponding shield operation parameters are adjusted according to the method used, and the corresponding parameter adjustment status will be transmitted to the jack hydraulic valve opening setter; the current jack partition grouping mode and jack switch status are read from the posture control auxiliary method library; When the predicted attitude of the shield machine meets the requirements of the ideal control target, it means that the current jack zone oil pressure distribution meets the requirements of reaching the ideal control target; when the predicted result meets the ideal control target, it means that the current jack zone oil pressure distribution method cannot meet the requirements of reaching the ideal control target. The evaluation method of the current control capability of the shield machine should consider the following factors: (1) The control deviation D of each control indicator i ; (2) Allowable error of each control index (3) The minimum tolerable index S of the overall fit.

2. The shield attitude precision control system facing the tunnel axis target according to claim 1 is characterized by: The working condition data acquisition module includes: a tunnel construction parameter and shield tunneling parameter acquirer and a current working condition feature extractor; The tunnel construction parameter and shield excavation parameter acquirer is used to extract the current relevant construction parameters of the shield from the shield sensing, measurement, and posture planning system, and the extracted information includes at least the following dimensions: cutter head torque, propulsion speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, upper / right / lower / left partition jack oil pressure valve opening, articulation level, articulation height, shield attitude control target, current shield attitude, current jack grouping status, and current jack activation status; the tunnel construction parameter and shield excavation parameter acquirer transmits the above parameters to the current working condition feature extractor and the shield attitude precise optimization control module; The current working condition feature extractor is used to extract the average value of the construction parameters of the shield in the latest construction ring excavation process from the tunnel construction parameters and the shield excavation parameter acquirer, which at least includes the following parameters: front soil pressure upper, front soil pressure right, front soil pressure lower, front soil pressure left, cutter head torque, cutter head speed, total thrust, oil pressure of the jacks in the upper / right / lower / left zones, and articulation horizontal degree and articulation height degree.

3. The shield attitude precision control system facing the tunnel axis target according to claim 1 is characterized by: The working condition feature recognition module includes a massive historical shield tunnel engineering construction data database and a working condition feature identifier; The massive historical shield tunnel engineering construction data database is used to store massive historical shield tunnel engineering construction data, wherein the data dimensions at least include cutter head torque, propulsion speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, articulation horizontal degree, and articulation height degree; The operating condition feature identifier is used to construct the operating condition feature identifier based on the operating condition feature data in a large amount of historical projects, and this process is established using a data-driven method or based on rules; The operating condition feature parameter type is consistent with the feature parameter type in the current operating condition feature extractor.

4. The shield attitude precision control system facing the tunnel axis target according to claim 1 is characterized by: The jack oil pressure valve opening setting device is used to read the current upper / right / lower / left partition jack oil pressure and upper / right / lower / left partition jack oil pressure valve opening information from the tunnel construction parameter and shield tunneling parameter acquisition device; The relative position relationship between the current shield attitude and the control target is read from the shield attitude control target; the jack oil pressure valve opening value is initialized according to the above information. During the operation of the system, the jack oil pressure valve opening setter reads the jack oil pressure recommended value output by the PSO attitude control parameter setting optimizer, updates the current setting value of the jack oil pressure valve opening of each partition of the shield, and adjusts the actual oil pressure value of each partition jack to the oil pressure recommended value; The jack oil pressure valve opening setting device outputs the zone jack oil pressure adjustment result to a massive historical shield tunnel engineering construction data database; The shield attitude optimization precise control module further includes a historical working condition feature identifier; the historical working condition feature identifier is used to cluster historical working condition feature data based on massive historical engineering data, and its working condition feature parameter type is consistent with the feature parameter type in the current working condition feature extractor; the K-Means algorithm is used to cluster the historical working conditions, and the number of working condition categories is determined according to the following method: Step 1: Assume that the initial number of working condition categories is k = 2, and k <n; Step 2: Calculate the Euclidean distance D between the sample data and the center point when the number of categories is k using the K-Means algorithm. k ; Step 3: Add one to the number of current working conditions, i.e. k+1; Step 4: Calculate the distance D between each sample and the center point after the number of working conditions increases by 1. k+1 , after the calculation category increases, the reduction in the Euclidean distance of the sample from the center point is G, that is, G = D k+1 -D k ; Step 5: Repeat Step 2 to Step 4. When G decreases, the k value at this time is the number of categories that the historical working conditions should be divided into; According to the above method, the number of categories determined by historical working conditions is determined, and their center points are numbered, which are used as working condition labels for historical construction data sections, and the working condition feature identifier is converted into a classifier function; Read the current working condition feature parameters passed in by the current working condition feature extractor, judge the category of the current working condition based on the Euclidean distance, take the working condition type with the smallest Euclidean distance from various types as the category of the current working condition, and output the location stamps of all sections in this category; The partitioned oil pressure posture relationship model is used to extract the jack partitioned oil pressure and the corresponding shield posture change of the corresponding section from the massive historical construction data based on the same working condition section position stamp output by the working condition feature identifier, and establish the relationship between the two based on the deep learning network; the input dimension of the network is the average oil pressure in the ε distance of the upper / right / lower / left partitioned jacks, expressed as The output dimension is the spatial variation of the shield cut and shield tail in the horizontal elevation direction and the variation of the elevation inclination angle. The distance span of the variation is consistent with the distance span ε for calculating the average oil pressure of the jack, which is expressed as Among them, the calculation formula of hidden layer neuron data is: Among them, h j is the calculation result of hidden layer neurons; w ij is the weight value between the input layer neurons and the hidden layer neurons; θ j is the bias value between the input layer neurons and the hidden layer neurons; f is the activation function set for the hidden layer; The zone oil pressure attitude relationship model calculates the spatial position change of the shield attitude driven by the current jack oil pressure at the ε distance span according to the oil pressure value of each zone of the shield jack under the current valve opening value input by the jack oil pressure valve opening setting device; The shield attitude control performance evaluator is used to evaluate the control effect based on the predicted value of the shield spatial position change at the ε distance and the ideal control target at the corresponding distance; The configuration of the attitude control parameter setting optimizer is as follows: (1) The constraints of the attitude control parameter setting optimizer include: ① the minimum thrust for driving the shield tunneling; ② the maximum number of jacks allowed to be closed in each zone; ③ the minimum oil pressure required for a single group of jacks to extend; ④ the number of jack zones; ⑤ the number of jacks in each zone; ⑥ the maximum oil pressure of the propulsion system oil pump; ⑦ the oil pressure valve opening control range; ⑧ the oil pressure valve opening-oil pressure value characteristic curve relationship; ⑨ the jack zone grouping method; (2) Evaluation function of the attitude control parameter setting optimizer: using the shield attitude control capability evaluator as the evaluation function in the optimization process; (3) The thrust calculation formula of the partition jack of the attitude control parameter setting optimizer is: F i =Num i *f i ; Among them, F i is the total pressure of the jack in zone i; Num i The number of jacks opened for partition i; f i The oil pressure of a single jack; (4) Relevant configuration is performed on the optimization algorithm of the attitude control parameter setting optimizer.

5. A shield attitude precision control method of the shield attitude precision control system facing the tunnel axis target according to any one of claims 1 to 4, characterized in that: The method comprises: The working condition data acquisition step is to obtain the real-time construction parameters of the shield machine, read the attitude control target at the current position, and extract the working condition characteristics of the current shield machine; The working condition feature recognition step is to establish a working condition identifier based on massive historical shield tunnel construction data using a data-driven or rule-based approach; The precise control steps of shield attitude optimization are to extract shield construction data under the same working conditions from massive historical construction data, and train to obtain the relationship model between the shield jack oil pressure distribution and attitude change; according to the ideal control target and shield attitude auxiliary method, the jack zone oil pressure distribution of the propulsion system is optimized to achieve precise control of the shield attitude.

6. The method for accurate control of shield attitude facing the tunnel axis target according to claim 5 is characterized in that: The operating condition data acquisition step comprises: The tunnel construction parameter and shield excavation parameter acquirer extracts the current relevant construction parameters of the shield from the shield sensor system, and the extracted information includes at least one of the following dimensions: cutter head torque, propulsion speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, upper / right / lower / left partition jack oil pressure valve opening, and articulation degree; the tunnel construction parameter and shield excavation parameter acquirer transmits the above parameters to the current working condition feature extractor and the jack oil pressure valve opening setter; The shield attitude control target building module obtains the control target of the current shield attitude from the shield trajectory planning system, that is, the expected spatial change of the shield cutout and the shield tail in the horizontal elevation direction, as well as the change of the elevation tilt angle; and transmits the control target to the jack hydraulic valve opening setter; The current working condition feature extractor extracts the mean value of the construction parameters of the shield during the latest construction ring excavation process from the tunnel construction parameters and the shield excavation parameter acquirer, which at least includes the following parameters: front earth pressure upper, front earth pressure right, front earth pressure lower, front earth pressure left, cutter head torque, cutter head speed, total thrust, oil pressure of the jacks in each upper / right / lower / left zone, and degree of articulation.

7. The shield machine posture precision control method facing the tunnel axis target according to claim 5 is characterized by: The operating condition feature identification step comprises: Stores a large amount of historical shield tunnel construction data, where the data dimensions at least include cutter head torque, propulsion speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, articulation horizontal degree, and articulation height degree; The operating condition feature identifier is constructed based on the operating condition feature data in massive historical projects. This process is established using a data-driven method or based on rules. The operating condition feature parameter type is consistent with the feature parameter type in the current operating condition feature extractor.

8. The method for accurate control of shield attitude facing the tunnel axis target according to claim 5 is characterized in that: The shield attitude optimization and precise control step comprises: A massive amount of historical shield tunnel construction data is stored in a massive historical shield tunnel construction data database, where the data dimensions of the tunnel construction data at least include cutter head torque, advancement speed, upper / right / lower / left soil pressure, total thrust, cutter head speed, upper / right / lower / left partition jack oil pressure, and articulation degree; The historical working condition feature identifier clusters the excavation working conditions according to the historical construction parameters; The zone oil pressure attitude relationship model fits the relationship between the zone oil pressure of the jack and the change of shield attitude under different working conditions according to historical data; The jack hydraulic valve opening setting device adjusts the corresponding hydraulic valve opening value according to the ideal hydraulic pressure of each partition, and transmits the value to the hydraulic valve actuator; The shield attitude control capability evaluator evaluates whether the current jack zone oil pressure can control the shield to reach the ideal control target, and evaluates its control effect, and cooperates with the attitude control parameter setting optimizer to optimize the zone oil pressure distribution; The attitude control parameter setting optimizer optimizes the oil pressure distribution of the jack zones in combination with the evaluation results of the shield attitude control capability evaluator, and hierarchically triggers the methods in the attitude control auxiliary method library according to the search results of the optimal solution; The posture control auxiliary method library provides at least two posture control auxiliary methods. When triggered and enabled, it adjusts the corresponding construction parameters and assists the posture control parameter setting optimizer to find the optimal solution for achieving precise control of the shield posture.

9. The method for precise control of shield attitude towards the tunnel axis target according to claim 8 is characterized in that: The posture control auxiliary method library includes the adjustment method of jack grouping, the jack switch control method, and the adjustment method of shield articulation degree; when the shield posture control auxiliary method is used, the corresponding shield operation parameters are adjusted according to the method used, and the corresponding parameter adjustment situation will be transmitted to the jack hydraulic valve opening setter; The jack oil pressure valve opening setting device reads the current upper / right / lower / left partition jack oil pressure and upper / right / lower / left partition jack oil pressure valve opening information from the tunnel construction parameter and shield tunneling parameter acquisition device; Read the relative position relationship between the current shield attitude and the control target from the shield attitude control target; read the current jack partition grouping method and the jack switch status from the attitude control auxiliary method library; initialize the jack oil pressure valve opening value according to the above information. During the system operation, the jack oil pressure valve opening setter reads the jack oil pressure recommended value output by the PSO attitude control parameter setting optimizer, updates the current setting value of the jack oil pressure valve opening of each shield partition, and adjusts the actual oil pressure value of each partition jack to the oil pressure recommended value; The jack oil pressure valve opening setting device outputs the zone jack oil pressure adjustment result to a massive historical shield tunnel engineering construction data database; The shield attitude optimization and precise control step includes a historical working condition feature extraction step, including: clustering historical working condition feature data based on massive historical engineering data by a historical working condition feature extractor, wherein the working condition feature parameter type is consistent with the feature parameter type in the current working condition feature extractor; clustering the historical working conditions by using a K-Means algorithm, and the number of working condition categories is determined according to the following method: Step 1: Assume that the initial number of working condition categories is k = 2, and k <n; Step 2: Calculate the Euclidean distance D between the sample data and the center point when the number of categories is k using the K-Means algorithm. k ; Step 3: Add one to the number of current working conditions, i.e. k+1; Step 4: Calculate the distance D between each sample and the center point after the number of working conditions increases by 1. k+1 , after the calculation category increases, the reduction in the Euclidean distance of the sample from the center point is G, that is, G = D k+1 -D k ; Step 5: Repeat Step 2 to Step 4. When G decreases, the k value at this time is the number of categories that the historical working conditions should be divided into; According to the above method, the number of categories determined by historical working conditions is determined, and their center points are numbered, which are used as working condition labels for historical construction data sections, and the working condition feature identifier is converted into a classifier function; Read the current working condition feature parameters passed in by the current working condition feature extractor, judge the category of the current working condition based on the Euclidean distance, take the working condition type with the smallest Euclidean distance from various types as the category of the current working condition, and output the location stamps of all sections in this category; The partitioned oil pressure posture relationship model is based on the same working condition section position stamp output by the working condition feature identifier, extracts the corresponding section jack partitioned oil pressure and the corresponding shield posture change from the massive historical construction data, and establishes the relationship between the two based on the deep learning network; the input dimension of the network is the average oil pressure in the ε distance of the upper / right / lower / left partitioned jacks, expressed as The output dimension is the spatial variation of the shield cut and shield tail in the horizontal elevation direction and the variation of the elevation inclination angle. The distance span of the variation is consistent with the distance span ε for calculating the average oil pressure of the jack, which is expressed as Among them, the calculation formula of hidden layer neuron data is: Among them, h j is the calculation result of hidden layer neurons; w ij is the weight value between the input layer neurons and the hidden layer neurons; θ j is the bias value between the input layer neurons and the hidden layer neurons; f is the activation function set for the hidden layer; The zone oil pressure attitude relationship model calculates the spatial position change of the shield attitude driven by the current jack oil pressure at the ε distance span according to the oil pressure value of each zone of the shield jack under the current valve opening value input by the jack oil pressure valve opening setting device; The shield attitude control performance evaluator evaluates the control effect based on the predicted value of the shield spatial position change at the ε distance and the ideal control target at the corresponding distance; When the predicted attitude of the shield machine meets the requirements of the ideal control target, it means that the current jack zone oil pressure distribution meets the requirements of reaching the ideal control target; when the predicted result meets the ideal control target, it means that the current jack zone oil pressure distribution method cannot meet the requirements of reaching the ideal control target. The evaluation method of the current control capability of the shield machine should consider the following factors: (1) The control deviation D of each control indicator i ; (2) Allowable error of each control index (3) the lowest tolerable index S of overall fit; The attitude control parameter setting optimizer adjusts the oil pressure valve opening of the jack oil pressure valve opening setter in combination with the evaluation result of the shield attitude control capability evaluator, thereby adjusting the oil pressure of each partition jack; when the attitude control parameter setting optimizer cannot find a feasible solution that meets the minimum index S of the shield attitude control capability evaluator by adjusting the oil pressure valve opening only, the relevant construction parameters of the current shield excavation are adjusted through the attitude control auxiliary method library; the attitude control parameter setting optimizer searches for the optimal jack partition oil pressure value distribution method according to the adjusted construction parameter configuration and the oil pressure attitude relationship model; The configuration of the attitude control parameter setting optimizer is as follows: (1) The constraints of the attitude control parameter setting optimizer include: ① the minimum thrust for driving the shield tunneling; ② the maximum number of jacks allowed to be closed in each zone; ③ the minimum oil pressure required for a single group of jacks to extend; ④ the number of jack zones; ⑤ the number of jacks in each zone; ⑥ the maximum oil pressure of the propulsion system oil pump; ⑦ the oil pressure valve opening control range; ⑧ the oil pressure valve opening-oil pressure value characteristic curve relationship; ⑨ the jack zone grouping method; (2) Evaluation function of the attitude control parameter setting optimizer: using the shield attitude control capability evaluator as the evaluation function in the optimization process; (3) The thrust calculation formula of the partition jack of the attitude control parameter setting optimizer is: F i =Num i *f i ; Among them, F i is the total pressure of the jack in zone i; Num i The number of jacks opened for partition i; f i The oil pressure of a single jack; (4) Relevant configuration is performed on the optimization algorithm of the attitude control parameter setting optimizer.

Images